Corporate failure diagnosis in SMEs. A longitudinal analysis based on alternative prediction models

• DOI: https://doi.org/10.1108/IJAIM-01-2013-0001
• Date: 25 February 2014
• Published date: 25 February 2014
• Pages: 49-67
• Author: Kosmas Kosmidis,Antonios Stavropoulos

Corporate failure diagnosis

in SMEs

A longitudinal analysis based on alternative

prediction models

Kosmas Kosmidis

Department of Information Management, Kavala Institute of Technology,

Kavala, Greece, and

Antonios Stavropoulos

Department of Applied Informatics, University of Macedonia,

Thessaloniki, Greece

Abstract

Purpose – The main purposes of this paper are to provide evidence about corporate failure diagnosis

in SMEs, identify the predictor variables that enhance the accuracy of the corporate failure diagnosis

models, and perform comparative analysis of the proposed models with the existing literature.

The paper supports the proposition that the majority of the proposed corporate failure diagnosis

models in the literature exhibit an endogenous drawback since their construction is based on large

entities or listed corporations’ samples.

Design/methodology/approach – The present study employs multiple discriminant analysis,

logit analysis, and probit analysis to construct corporate failure diagnosis models based on SMEs

longitudinal data from Greece.

Findings – The paper provides evidence that the contribution of human capital is immensely more

important to the viability of SMEs than to the viability of large corporations. Moreover, this study

identif‌ies interactions among seemingly insignif‌icant variables that exhibit incremental information

content and attribute massive discriminant power to the proposed corporate failure diagnosis models.

Practical implications – The results of this study encourage regulatory authorities to adopt

enhancements to the Basel II framework and f‌inancial institutions as regards to constructing their

corporate failure diagnosis models. The models is based upon internal default experience and mapping

to external data incorporating both quantitative and qualitative variables.

Originality/value – The contribution of this paper is the proposition of new value-relevant variables

that enhance the accuracy of existing corporate failure diagnosis models for SMEs.

Keywords Greece, SMEs,Finance, Bankruptcy, Corporate failure, Prediction models

Paper type Research paper

1. Introduction

The diagnosis of corporate failure has been the apple of discord among researchers,

academics and professionals for the last four decades since the pioneering work of

Altman (1968), who employed multiple discriminant analysis methodology (MDA) in

order to predict corporate bankruptcy. Apparently, this debate is becoming timelier

nowadays due to the rampant spread of f‌inancial turbulence, which reinforces the

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1834-7649.htm

The authors are grateful to Konstantinos Terzidis, Evangelos Tsoukatos and two anonymous

referees for their useful comments and suggestions.

Received 14 January 2013

Revised 14 January 2013

Accepted 14 January 2013

International Journal of Accounting

and Information Management

Vol. 22 No. 1, 2014

pp. 49-67

qEmerald Group Publishing Limited

1834-7649

DOI 10.1108/IJAIM-01-2013-0001

Corporate failure

diagnosis in

SMEs

49

proclaimed important role of corporate failure diagnosis. The prime causes of failure in

mature companies are concerned with fatal corporate strategy decisions and especially

with “defective response to change” (Argenti, 1976). Therefore, ability to adapt to a

ceaselessly changing business environment is the cornerstone of a f‌irm’s potential to

survive in the global arena.

Corporate failure affects a plethora of stakeholders such as employees, managers,

shareholders, auditors, creditors, and, to the extent that failure results in breaking up a

corporation’s social and economic interaction with its host environment, the society as

a whole. The devastating impact that the collapse of Enron, Worldcom, Barings Bank,

Imarbank and others had on the Global economy supports the preceding argument

about the plethora of interested parties affected by corporate failure. Numerous studies

on f‌inancial distress signalling and corporate failure prediction have been reported in

the literature. However, in their vast majority, these are conf‌ined to large entities or

listed corporations.

Nevertheless, the birth of the current economic crisis was not the collapse of few

colossal corporations but the massive default of US households in the sub-prime

mortgage credit market. Consequently, research on corporate failure diagnosis

should be expanded to small- and medium-sized enterprises (SMEs) and non-listed

corporations, due to their large number and their impact on real economy. Especially in

Europe, the majority of enterprises are considered small and account for a signi f‌icant

amount of European work experience and economic activity (Baixauli and

Modica-Milo, 2010). In particular, this study focuses on Greece, which exhibits many

similarities with other Southern European countries concerning the volume and

frequency rate of SMEs in the economy (Commission of the European Union, 2007).

The main objectives of this paper are to:

.provide evidence on corporate failure diagnosis in SMEs;

.identify the f‌inancial ratios enhancing the predictive ability of corporate failure

diagnosis models; and

.perform a comparative analysis of the proposed models in relation to existing

literature.

The main contribution of this paper is to propose new value-relevant variables that

improve the accuracy of existing models for SMEs. Evidence on bankrupt corporations

was collected from the county courts in the region of Central and Eastern Macedonia

in Greece.

The remainder of this paper is organised as follows: Section 2 presents an extensive

literature review concerning corporate failure diagnosis and the current advances with

the employment of experts systems. Our research methodology including sample and

variable selection processes is provided in Section 3 of this paper. Section 4 reports the

empirical results of our analysis. The discussion of the results and the practical

implications of our study are embedded in Section 5. Finally, our concluding remarks

are cited in Section 6 of this paper.

2. Literature review

There is a plethora of studies concerning f‌inancial distress signalling and corporat e

failure diagnosis. The ability to discriminate between f‌inancially distressed and viable

corporations was enhanced by the use of f‌inancial ratios. The pioneering work of

IJAIM

22,1

50